On 08/15/2015 10:54 AM, Martin Kletzander wrote: > On Sat, Aug 15, 2015 at 09:18:14AM -0400, John Ferlan wrote: >> Coverity complained that 'vm' wasn't initialized before jumping to >> cleanup: and calling virDomainObjEndAPI if the VIR_STRDUP fails. >> Rather than initialize vm = NULL, I moved the VIR_STRDUP closer to >> usage and used endjob for goto. >> > > Wild pointers should not exist. At all. It's very subjective, but I > just don't like them. ACK if you also initialize that pointer to > NULL; that won't hurt anyone and will avoid possible future problems. > I agree in general - although in qemu_driver.c - it's hit or miss whether 'vm' is initialized to NULL. I just flipped a coin and it landed on moving the code instead of initializing the pointer ;-) In any case added the = NULL; and pushed. John -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list